diff options
| author | Bob Peterson <rpeterso@redhat.com> | 2023-03-23 19:26:02 +0300 |
|---|---|---|
| committer | Andreas Gruenbacher <agruenba@redhat.com> | 2023-03-23 21:37:56 +0300 |
| commit | 260595b439776c473cc248f0de63fe78d964d849 (patch) | |
| tree | ca4e0934d7bde14db9323cceaadb78ce614cdc52 | |
| parent | e8d018dd0257f744ca50a729e3d042cf2ec9da65 (diff) | |
| download | linux-260595b439776c473cc248f0de63fe78d964d849.tar.xz | |
Reinstate "GFS2: free disk inode which is deleted by remote node -V2"
It turns out that reverting commit 970343cd4904 ("GFS2: free disk inode
which is deleted by remote node -V2") causes a regression related to
evicting inodes that were unlinked on a different cluster node.
We could also have simply added a call to d_mark_dontcache() to function
gfs2_try_evict(), but the original pre-revert code is better tested and
proven.
This reverts commit 445cb1277e10d7e19b631ef8a64aa3f055df377d.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
| -rw-r--r-- | fs/gfs2/dentry.c | 18 |
1 files changed, 18 insertions, 0 deletions
diff --git a/fs/gfs2/dentry.c b/fs/gfs2/dentry.c index 6fe9ca253b70..2e215e8c3c88 100644 --- a/fs/gfs2/dentry.c +++ b/fs/gfs2/dentry.c @@ -83,8 +83,26 @@ static int gfs2_dhash(const struct dentry *dentry, struct qstr *str) return 0; } +static int gfs2_dentry_delete(const struct dentry *dentry) +{ + struct gfs2_inode *ginode; + + if (d_really_is_negative(dentry)) + return 0; + + ginode = GFS2_I(d_inode(dentry)); + if (!gfs2_holder_initialized(&ginode->i_iopen_gh)) + return 0; + + if (test_bit(GLF_DEMOTE, &ginode->i_iopen_gh.gh_gl->gl_flags)) + return 1; + + return 0; +} + const struct dentry_operations gfs2_dops = { .d_revalidate = gfs2_drevalidate, .d_hash = gfs2_dhash, + .d_delete = gfs2_dentry_delete, }; |